This invention relates generally to processing raw data and, more particularly, to computer systems and computer-based methods for creating standardized street addresses from raw address data.
There are several forms of raw data associated with a payment-by-card transaction. Such raw data may be provided to and from a merchant, a merchant bank, a merchant processor, or other party involved in the payment-by-card transaction. The raw data may include a transaction identification number, a merchant identification number, an authorization code, address data, etc. Prior to authorizing and/or settling the transaction, one or more portions of the raw data is generally reconciled with one or more entries in a database to, for example, match portions of the merchant's raw data with data contained in a merchant database. The merchant database may be a repository of merchant specific data maintain by a merchant bank, issuer, or a merchant processor. Additionally, one or more parties may provide raw data unassociated with a transaction for reconciliation with data in the merchant database.
In processing the raw data from one or more merchants, the raw address data is often received in a variety of different formats, possibly depending on the merchant, country of origin, or type of merchant, etc. Raw street addresses, for example, may include special characters, a street number in different locations, directional indicators, apartment indicators, extraneous information, etc. Different matching methods have been employed to match raw address data to address data contained in a merchant database. In particular, known methods include term frequency/inverse document weighting, edit distances, and soundex for measuring similarities between two strings or records. Unfortunately, each of these known methods of matching data has limitations.
Accordingly, it would be desirable to provide a system and/or method for creating a standardized street address from a raw street address for comparison to data stored in a database.